Multi-Scale Model of the U.S. Transportation Energy Market for Policy Assessment

Eppstein, Margaret J.; Rizzo, Donna M.; Marshall, Jeffrey S. · 2013 · ROSA P / University of Vermont. Transportation Research Center

archive: archived pipeline: cataloged verified

Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)

Summary

This report presents a multi-scale modeling framework designed to assess Plug-in Hybrid Electric Vehicle (PHEV) market penetration and inform policy decisions. The research addresses the challenge that transportation is the fastest-growing source of U.S. greenhouse gas emissions, yet PHEV adoption is hindered by consumer unfamiliarity, range anxiety, and complex interactions between financial incentives, social influences, and regulatory environments. The authors aimed to create a model capable of simulating these nonlinear dynamics, up-scaling results to national levels, and identifying the legal and regulatory changes necessary for widespread adoption. The methodology combines an Agent-Based Model (ABM) with Artificial Neural Networks (ANNs) and legal analysis. The ABM simulates individual consumer agents with attributes such as income, vehicle miles traveled, and susceptibility to social influence. Agents make vehicle choices based on rational cost estimates and heuristic factors, such as environmental concern, which are dynamically influenced by media coverage and peer networks. To overcome the computational limits of the ABM at large scales, the researchers developed a Generalized Regression Neural Network (GRNN) trained on ABM outputs to act as a fast function approximator for regional simulations. Additionally, the team assembled a dataset of 321,487 Vermont passenger vehicles to analyze spatial fuel efficiency patterns and developed a novel Cluster Reinforcement method for Self-Organizing Maps to visualize high-dimensional data. A legal case study with the Vermont Law School examined regulatory barriers, and a consumer survey via Amazon Mechanical Turk gathered data on adoption thresholds. Key findings indicate that providing consumers with accessible lifetime fuel cost estimates is critical for promoting PHEV adoption. The model suggests that increasing gasoline prices can nonlinearly magnify market penetration, whereas temporary tax credits are ineffective for long-term fleet efficiency unless accompanied by reductions in vehicle sticker prices. Longer battery ranges were identified as a significant leverage point that amplifies the impact of price reductions. The GRNN successfully replicated ABM behavior with high accuracy, reducing simulation time from hours to seconds, thereby enabling large-scale policy testing. The legal analysis recommended that Vermont enact specific tax incentives, streamline public charging infrastructure regulations, integrate PHEVs into utility Integrated Resource Plans, and adopt uniform off-peak charging rates to incentivize grid-friendly behavior. The significance of this work lies in its integrated approach to modeling complex socio-technical systems. By combining micro-level behavioral modeling with macro-level computational scaling and legal analysis, the study provides policymakers with tools to evaluate the synergistic effects of fuel taxes, battery technology improvements, and regulatory frameworks. The development of the GRNN up-scaling method and the Cluster Reinforcement visualization technique also contributes to the broader field of computational engineering, offering efficient methods for analyzing large-scale discrete-choice models and high-dimensional datasets.

Key finding

Providing consumers with readily accessible estimates of lifetime vehicle fuel costs and increasing gasoline prices can non-linearly magnify PHEV market penetration, whereas temporary tax credits are unlikely to have lasting effects on long-term fleet efficiency unless manufacturers lower sticker prices after rebates end.

Methodology

mixed_methods

Sample size: 1000

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• acceptance adoption